Skip to main content

Rectoneovaginal Fistula (RNVF) Repair

Rectoneovaginal fistula (RNVF) — an abnormal epithelialised communication between the rectum and the neovaginal canal — is one of the most clinically significant major complications after feminizing genital surgery, occurring in 0.4–2.3% of primary vaginoplasties and up to 6.3% of revision vaginoplasties.[1][2][3][4] The largest published experience (Amsterdam, van der Sluis 2016, n = 1,082 vaginoplasties with 13 RNVFs) demonstrates that fistulectomy + primary closure or local advancement flap suffices in most patients, with few complications and no impairment of neovaginal function.[1]

This is the dedicated management-pathway page. For the host operations, see Penile Inversion Vaginoplasty, Peritoneal Pull-Through Vaginoplasty, and Intestinal Vaginoplasty. For the related stenosis-management pathway, see Neovaginal Stenosis Management.

Definitions and Classification​

Types of neovaginal fistulas after vaginoplasty:[1]

  • Rectoneovaginal fistula (RNVF) — rectum to canal; flatus / stool / feculent discharge through the neovagina (most clinically significant).
  • Urethroneovaginal fistula (UNVF) — urethra to canal; continuous urinary leakage.
  • Pouch-neovaginal fistula — canal to peritoneal pouch (rare; 1/1,082 in Amsterdam).
  • Prostato-vaginal fistula — prostatic urethra / prostate to canal (4 patients underwent robot-assisted sigmoid vaginoplasty for this in the Sljivich series).[5]

By location (adapted from general RVF classification):[6][7]

  • Low — at or near introitus / perineal body; most accessible for perineal repair.
  • Mid-canal — within the canal body (most common in GAS context; mean 5.4 cm from introitus in the Pansritum series).[8]
  • High / apical — near apex; may require transabdominal approach.

By complexity:[6]

  • Simple — small (< 1 cm), single tract, healthy surrounding tissue.
  • Complex — > 1 cm, multiple tracts, significant scarring, recurrent after prior repair, concurrent stenosis.

Incidence by Surgical Setting​

SeriesnRNVF rateNotes
PIV — Amsterdam (Buncamper)[4]4750.6%Median follow-up 7.8 yr
PIV — Gaither[3]3300.9%—
PIV — Levy[9]2400.5%—
PIV — OHSU (anatomy-guided dissection)[2]2001.0%—
PIV — Stark[10]1,0110.1%Only patient with prior perineal surgery; no intraoperative sigmoidoscopy
All primary vaginoplasty — Amsterdam[1]1,0370.8%—
Revision vaginoplasty — Amsterdam[1]806.3%OR 8.6 vs primary, p < 0.01
Robotic peritoneal flap (NYU Xi vs SP)~ 5001%—
Robot-assisted sigmoid (Sljivich)[5]360% new fistulas4 patients underwent RSV for pre-existing fistula
Pansritum 2-centre Thai cohort[8]2,0590.24% (5/2,059)2 from external referrals

Revision vaginoplasty carries an 8.6× increased risk vs primary — the dominant procedural risk factor — attributable to scarred fibrotic tissue planes that obscure normal anatomy.[1]

Pathophysiology and Risk Factors​

Mechanism​

The neovaginal canal is dissected in the rectoprostatic / rectovesical space, with the rectum at greatest risk because the canal lies directly anterior to the anterior rectal wall, separated only by Denonvilliers' fascia.[2][11] RNVF develops through two primary mechanisms:[1][10]

  1. Intraoperative rectal injury — direct perforation either unrecognised or failing to heal despite primary repair.
  2. Postoperative ischemic necrosis — devascularisation of the rectoneovaginal septum with delayed tissue breakdown.

Risk factors​

  • Revision vaginoplasty — OR 8.6 for RNVF vs primary (p < 0.01).[1]
  • Intraoperative rectal perforation — 4/23 (17.4%) of intraoperatively recognised and oversewn perforations subsequently developed an RNVF (Amsterdam) — recognised injuries still carry significant fistula risk.[1]
  • Prior perineal surgery — only patient with RNVF in the Stark 1,011-patient series.[10]
  • Failure to perform intraoperative sigmoidoscopy with air-leak test after rectal injury.[10]
  • Diabetes — local-infection OR 9.8 (p = 0.003) which may predispose to fistula.[2]
  • Concurrent vaginal stenosis — 50% of patients with rectal injury who proceed with full-depth GAV develop stenosis, which may contribute to fistula via tissue ischemia.[10]
  • Smoking — only negative predictive factor for fistula-repair success in the Pastier multicentre study (p = 0.02).[12]

Clinical Presentation and Diagnosis​

Symptoms[1][6]​

  • Passage of flatus through the neovagina (earliest and most specific).
  • Stool or feculent discharge through the neovagina.
  • Foul-smelling discharge.
  • Recurrent UTI.
  • Perineal pain.
  • Difficulty / pain / bleeding during dilation.

Diagnostic workup​

In the Amsterdam experience, most patients are diagnosed on symptoms and exam alone.[1] Systematic evaluation:[7][13][14]

  • Neovaginal speculum exam — direct fistula visualisation on the posterior canal wall; feculent material or bubbling.
  • Digital rectal exam — palpate the tract on the anterior rectal wall.
  • MRI pelvis — modality of choice; 87% accuracy alone, 100% combined with EUA.[7][13]
  • Endoanal / endorectal ultrasound (EAUS) — 91% accuracy; assesses tract and sphincter integrity.[7]
  • Examination under anaesthesia (EUA) — 91% accuracy; direct visualisation + hydrogen-peroxide instillation.[7][15]
  • Sigmoidoscopy / proctoscopy — rectal side + concurrent pathology.
  • Fluoroscopic or CT vaginography — for upper / mid-canal fistulas.[14]
  • Combining any two modalities achieves 100% diagnostic accuracy.[7]

Prevention — Intraoperative Strategies​

Anatomically guided dissection (OHSU / Shoureshi–Dugi)​

200-patient series with 1.0% full-thickness rectal injury and 1.0% fistula rate:[2]

  • Dissection begins at the central tendon of the perineum.
  • Plane developed between rectum and bulbospongiosus / bulbar urethra.
  • Denonvilliers' fascia identified as the landmark to maintain correct plane.
  • Cephalad dissection in the avascular rectoprostatic plane.
  • Finger in the rectum for continuous tactile feedback.

Morris algorithmic approach to intraoperative rectal injury (n = 146; 6.8% rectal-injury rate)[16]​

  1. Mechanical bowel prep in all patients.
  2. Continuous rectal-finger palpation during dissection; immediate injury recognition.
  3. If rectal injury identified:
    • Immediate multilayer repair (two-layer most; three-layer for larger injuries).
    • Intraoperative CRS consultation → sigmoidoscopy with air-leak test.[10][16]
    • Bulbospongiosus muscle interposition between rectal repair and canal lining (2/9 patients in Stark).[10]
    • Convert decision: secure repair + negative air leak → proceed with full-depth (counsel re ~50% stenosis risk); insecure / extensive → convert to minimal-depth (zero-depth) vaginoplasty.[10]
    • Temporary fecal diversion for large injuries / insecure repairs (2/9 in Stark).[10]
  4. Key Stark finding: when injuries are identified intraoperatively, multilayer-repaired, and CRS-evaluated, patients do not develop RVF despite completing full-depth GAV — the single RVF in the series had prior perineal surgery and no intraoperative sigmoidoscopy.[10]

Rectoprostatic fascia reinforcement flap (Pansritum)[8]​

Novel preventive / repair technique using the rectoprostatic (prostaticovesicular) fascia as a vascularised flap interposed between rectal repair and canal lining. 2,059 vaginoplasties → 14 intraoperative rectal injuries + 5 RNVFs treated; mean location 5.2 cm injuries / 5.4 cm fistulas from introitus; mean size 2.5 / 2.3 cm; 0% recurrence at 1 yr. Applicable to both intraoperative repair and delayed RNVF; uses in-situ AMAB-specific tissue, no distant donor site.

Stepwise Management Algorithm​

Step 1 — Assessment and optimisation[1][6][7]​

  • Define anatomy (MRI + EUA).
  • Identify concurrent pathology (stenosis, granulation, infection, abscess).
  • Optimise tissue (resolve inflammation, treat infection, nutrition).
  • Smoking cessation — the only negative predictor of repair success (p = 0.02).[12]
  • Timing — wait 3–6 mo from primary surgery or fistula onset to allow inflammation to settle and tissues to mature.[6][7]

Step 2 — Conservative management​

For small, minimally symptomatic fistulas:[7]

  • Wound care, sitz baths, fibre supplements.
  • Antibiotics if infection.
  • Cessation of dilation until repair.
  • 52–66% non-operative healing reported for benign RVFs in general populations (limited GAS-specific data).

Step 3 — Seton placement (selected cases)​

Per ASCRS 2022 (Grade 1C):[7] draining seton for narrow fistulas, small vaginal-side openings, multiple tracts, or to settle acute inflammation / infection prior to definitive repair.

Step 4 — Definitive repair​

A. Fistulectomy + primary closure ± local advancement flap (first-line) — Amsterdam first-choice; tract excision + multilayer closure (rectal layers → interposition → canal wall).[1]

B. Endorectal advancement flap (ERAF) — ASCRS-preferred procedure for most RVFs (Grade 1C). Success 41–78%; failure factors: sphincter abnormalities, prior radiation, recurrence. Repeat flaps 55–93% success. Fistula diameter > 1 cm is an independent ERAF-failure predictor (p < 0.05).[7][17]

C. Rectoprostatic fascia reinforcement flap (Pansritum) — 100% success (5/5) at 1 yr; in-situ AMAB-specific tissue.[8]

D. Martius (bulbocavernosus) flap — typically recommended for recurrent or complex RNVF (ASCRS Grade 1C). Success 65–94%. Pastier multicentre n = 62: Martius = gracilis on success (69% vs 69%) but Martius had shorter LOS and 27% performed without a stoma with no morbidity impact. In GAS, the bulbospongiosus muscle and surrounding labia-majora fat pad are harvested from the reconstructed labia majora and tunneled subcutaneously to the fistula site.[7][12][10][18]

E. Gracilis muscle interposition — salvage after Martius failure or for complex / recurrent fistulas. Maspero SR pooled success 64% (95% CI 53–74%) across 384 patients. Failure factors: smoking, Crohn's, > 2 prior repairs. Stoma associated with improved outcomes. Postop complications 0–37% (SSI, thigh numbness, hematoma). Longer LOS than Martius (+ 2 d, p = 0.01). Success of gracilis after failed Martius not significantly different from initial gracilis (p > 0.99) — reserve as salvage.[12][19]

F. Robot-assisted sigmoid vaginoplasty for concurrent fistula + stenosis — Sljivich n = 4 prostato-vaginal-fistula patients underwent RSV; 0% new fistulas; mean depth 17.6 cm. Particularly useful when the canal is severely stenosed or obliterated and the fistula cannot be accessed for local repair.[5] Cross-link to the intestinal vaginoplasty page sigmoid-revision section.

G. Robotic transanal minimally invasive surgery (R-TAMIS) — emerging; da Vinci Xi via GelPOINT Path; vaginal wall closed with barbed sutures reinforced with fibrin sealant + acellular dermal mesh; same-day discharge reported. Currently case-report level; not specifically studied in neovaginal context.[20]

H. Transabdominal salvage (last resort) — completion proctectomy with coloanal anastomosis (ASCRS Grade 2C) or Turnbull–Cutait staged coloanal anastomosis for the most refractory cases.[7]

Fecal Diversion​

Nuanced and individualised role:[1][7][12][21]

Amsterdam experience (GAS-specific):[1]

  • 4 of 13 RNVF patients required diversion — 1 with concurrent closure, 3 staged (diversion → delayed closure).
  • Remaining 9 patients repaired without diversion.

General RVF evidence:[12][21][7]

  • Obi 2026 (n = 158; 424 procedures) — multiple prior repairs and interposition-flap repairs associated with diversion use (p < 0.05).[21]
  • Pastier 2024 — 27% of Martius flap repairs performed without stoma with no morbidity impact (p = 0.763).[12]
  • ASCRS 2022 — diverting stoma not shown to improve ERAF outcomes; individualised; generally recommended adjunct to Martius and gracilis flaps despite weak evidence.[7]

Practical guidance:

  • Small / simple RNVF → repair without diversion is reasonable.
  • Large / complex / recurrent → strongly consider diversion.
  • Concurrent stenosis → diversion may be needed for staged repair (fistula → stenosis).
  • Stoma closure typically within 3 mo after confirming healing.[7]

Comparison of Repair Techniques​

TechniqueSuccessStoma?Best indication
Fistulectomy + primary closureHigh (Amsterdam)Usually noSmall, simple, first-time RNVF[1]
ERAF41–78%Usually noSimple-to-moderate RNVF[7]
Rectoprostatic fascia flap100% (5/5)NoIntraoperative injury or delayed RNVF; AMAB-specific in-situ tissue[8]
Martius flap65–94%Optional (27% without)Recurrent or complex RNVF[7][12]
Gracilis interposition64% (pooled)Usually yesSalvage after Martius failure[12][19]
Robot-assisted sigmoid vaginoplasty100% (4/4)NoRNVF + severe stenosis / obliteration[5]
R-TAMISCase reportsNoEmerging; high RVF in selected cases[20]

Outcomes and Prognosis​

Reassuring Amsterdam framing — surgical RNVF repair is associated with few intraoperative and postoperative complications and does not impair neovaginal function.[1] Almost all patients can ultimately be closed, though recurrences are common and multiple procedures may be required.[6]

Factors associated with success (general RVF, applicable to RNVF)[22][12][21]​

  • Major procedure (vs conservative) — OR 6.4 (p < 0.05).
  • Diverting stoma — OR 3.5 (p = 0.009) for success.
  • First surgery at experienced centre — OR 3.2 (p = 0.003).
  • Smoking — only negative predictor (p = 0.02).
  • Fistula diameter > 1 cm — independent ERAF-failure predictor.

Impact on neovaginal function​

Successful RNVF repair does not impair neovaginal function — patients maintain depth and the ability to dilate and engage in intercourse after repair. However, concurrent stenosis (in 50% of patients with rectal injury proceeding to full-depth GAV) requires separate management.[1][10]

Special Considerations​

  • Concurrent stenosis — staged approach (fistula closure ± diversion → stenosis management) or RSV addressing both simultaneously.[5][10] See Neovaginal Stenosis Management.
  • Multidisciplinary team — primary vaginoplasty surgeon + colorectal surgery + urogynecology. Stark emphasises intraoperative CRS consultation with sigmoidoscopy + air-leak testing when rectal injury identified.[10][15]
  • Dilation during fistula management — suspend or modify during active management to avoid mechanical disruption; resume gradually after closure. Creates tension between fistula healing and stenosis prevention.[1]
  • Recurrence — Corte 2015 (n = 79; 286 procedures) — aggressive treatment with early stoma + major procedure after local failure → overall 72% success; highest chance is the initial repair attempt, underscoring importance of thorough preoperative workup and optimal first-attempt technique.[22][15]

Evidence Limitations​

  • No GAS-specific guidelines — practice adapted from general RVF guidelines (ASCRS 2022) and limited case series.[1][7]
  • Very small sample sizes — largest GAS-specific RNVF series has 13 patients (Amsterdam).[1]
  • No comparative trials between repair techniques in the neovaginal context.
  • Unique anatomy — prostate retention, absence of native vagina, skin-/peritoneum-/bowel-lined canal — fundamentally different surgical environment than cisgender RVF.
  • Limited long-term follow-up — most series report 1-year outcomes.
  • Rectoprostatic fascia flap (Pansritum) is the only GAS-specific technique, based on only 5 fistula repairs.[8]
  • No standardised outcome measures — clinical healing vs radiographic closure vs PRO definitions vary.
  • Quality of evidence uniformly low — Göttgens 2014 SR characterised the published quality as "disappointing" with no RCTs and closure rates 0–> 80%.[23]

References​

1. van der Sluis WB, Bouman MB, Buncamper ME, et al. Clinical characteristics and management of neovaginal fistulas after vaginoplasty in transgender women. Obstet Gynecol. 2016;127(6):1118–1126. doi:10.1097/AOG.0000000000001421

2. Shoureshi P, Dy GW, Dugi D. Neovaginal canal dissection in gender-affirming vaginoplasty. J Urol. 2021;205(4):1110–1118. doi:10.1097/JU.0000000000001516

3. Gaither TW, Awad MA, Osterberg EC, et al. Postoperative complications following primary penile inversion vaginoplasty among 330 male-to-female transgender patients. J Urol. 2018;199(3):760–765. doi:10.1016/j.juro.2017.10.013

4. Buncamper ME, van der Sluis WB, van der Pas RSD, et al. Surgical outcome after penile inversion vaginoplasty: a retrospective study of 475 transgender women. Plast Reconstr Surg. 2016;138(5):999–1007. doi:10.1097/PRS.0000000000002684

5. Sljivich M, Torres C, Chen D, et al. Feasibility and outcomes after robot-assisted sigmoid vaginoplasty for gender dysphoria. Urology. 2025. doi:10.1016/j.urology.2025.06.003

6. Snyder MJ. Management of rectovaginal fistulae. Surg Clin North Am. 2026;106(1):65–76. doi:10.1016/j.suc.2025.08.004

7. Gaertner WB, Burgess PL, Davids JS, et al. The American Society of Colon and Rectal Surgeons clinical practice guidelines for the management of anorectal abscess, fistula-in-ano, and rectovaginal fistula. Dis Colon Rectum. 2022;65(8):964–985. doi:10.1097/DCR.0000000000002473

8. Pansritum K, Thomrongdullaphak S, Suwajo P. A rectoprostatic fascia reinforcement flap for rectal injury and rectoneovaginal fistula in gender-affirmation surgery. Plast Reconstr Surg. 2022;150(4):909–913. doi:10.1097/PRS.0000000000009575

9. Levy JA, Edwards DC, Cutruzzula-Dreher P, et al. Male-to-female gender reassignment surgery: an institutional analysis of outcomes, short-term complications, and risk factors for 240 patients undergoing penile-inversion vaginoplasty. Urology. 2019;131:228–233. doi:10.1016/j.urology.2019.03.043

10. Stark T, Celtik K, Ting J, Purohit RS. Rectal injury during penile inversion vaginoplasty. Urology. 2024;192:141–145. doi:10.1016/j.urology.2024.05.043

11. Morrison SD, Claes K, Morris MP, et al. Principles and outcomes of gender-affirming vaginoplasty. Nat Rev Urol. 2023;20(5):308–322. doi:10.1038/s41585-022-00705-y

12. Pastier C, Loriau J, Denost Q, et al. Rectovaginal fistula: what is the role of Martius flap and gracilis muscle interposition in the therapeutic strategy? Dis Colon Rectum. 2024;67(8):1056–1064. doi:10.1097/DCR.0000000000003148

13. VanBuren WM, Lightner AL, Kim ST, et al. Imaging and surgical management of anorectal vaginal fistulas. Radiographics. 2018;38(5):1385–1401. doi:10.1148/rg.2018170167

14. Expert Panel on Gastrointestinal Imaging, Levy AD, Liu PS, et al. ACR Appropriateness Criteria Anorectal Disease. J Am Coll Radiol. 2021;18(11S):S268–S282. doi:10.1016/j.jacr.2021.08.009

15. Muñoz JM, Levin PJ, Saur NM, Cox CK. Multidisciplinary approach to evaluation of rectovaginal fistulas. Am J Obstet Gynecol. 2024;231(2):285–286. doi:10.1016/j.ajog.2024.04.029

16. Morris MP, Wang CW, Holan C, et al. Rectal injury during penile inversion vaginoplasty: an algorithmic approach to prevention and management. Plast Reconstr Surg. 2023;152(2):326e–337e. doi:10.1097/PRS.0000000000010246

17. Li X, Shao W, Sun G. A single-center retrospective analysis of endorectal advancement flaps used for the treatment of simple rectovaginal fistulas. Scand J Gastroenterol. 2025;60(4):307–311. doi:10.1080/00365521.2025.2468493

18. Wang D, Chen J, Zhu L, et al. Surgical repair of rectovaginal fistula using the modified Martius procedure: a step-by-step guide. J Minim Invasive Gynecol. 2018;25(4):573–575. doi:10.1016/j.jmig.2017.08.657

19. Maspero M, Otero Piñeiro A, Steele SR, Hull TL. Gracilis muscle interposition for the treatment of rectovaginal fistula: a systematic review and pooled analysis. Dis Colon Rectum. 2023;66(5):631–645. doi:10.1097/DCR.0000000000002739

20. Mohammed Salih S, Zajicek J, Allawi A. Same-day repair of recto-vaginal fistula using the robotic trans-anal minimally invasive technique — how we do it. J Laparoendosc Adv Surg Tech A. 2025. doi:10.1177/10926429251399211

21. Obi M, Kanters A, Spivak AR, et al. Factors associated with fecal diversion prior to rectovaginal fistula repair. J Gastrointest Surg. 2026:102442. doi:10.1016/j.gassur.2026.102442

22. Corte H, Maggiori L, Treton X, et al. Rectovaginal fistula: what is the optimal strategy? An analysis of 79 patients undergoing 286 procedures. Ann Surg. 2015;262(5):855–860; discussion 860–861. doi:10.1097/SLA.0000000000001461

23. Göttgens KW, Smeets RR, Stassen LP, Beets G, Breukink SO. The disappointing quality of published studies on operative techniques for rectovaginal fistulas: a blueprint for a prospective multi-institutional study. Dis Colon Rectum. 2014;57(7):888–898. doi:10.1097/DCR.0000000000000147